In order to evaluate the pragmatic components of the definition of prescription, we might ask whether a prescription which has no authority at all over us is still a prescription?
For instance, would a prescription about the way ants should be cooked found in a book on anthropology of Central African populations still be a prescription if read by a vegetarian reader? Generally speaking, the answer depends on how one defines a prescription (formally or pragmatically). From the point of view of the vākyaśāstra (about which see here), I am inclined to think (although the example is not dealt with in texts I am aware of) that it would NOT be one, since an exhortation presupposes the listener's desire for the result to be achieved through the action enjoined. In the case of a slave being directed to carry something, the desire could be rephrased as her desire to satisfy her master, but the role of desire cannot be ruled out altogether. Hence, exhortation is not thought of as a purely linguistic phenomenon, independent of whoever listens to it. There is a link to pragmatics, via the definition of a context, of suitable listeners, and of the range of authority of the utterance (in the Speech Act Theory, this would be tantamount to the range of authority of the speaker).
A Sanskritist may now object that no Indian author would accept that the Vedas might be thought as not prescriptive. There is indeed a subtle boundary between what is ontologically not a prescription and what just does not function subjectively like one. In general, I am inclined to think that a vākyaśāstra-author would say that the Vedas do not cease to be prescriptive, because they address people who desire happiness, and everyone desires happiness. But I never found an explicit argument for it in the texts I am aware of.
On vākyaśāstra, see here. On the Speech-Act theory, see here and here.
On prescriptions, see also the corresponding tag on the left.
On the necessity of desire, see here.